Brain mechanisms of vulnerability to suicide in old age remain unclear, and very few researchers study the neurobiology of late-life'suicide. In particular, the field of suicide research lacks an understanding of how psychoiogicai and cognitive markers of suicidal risk (hopelessness, executive dyscontrol) relate to brain markers identified in post-mortem and imaging studies. The applicant - a geriatric psychiatrist with a clinical background in late-life suicide and in the treatment of late-life depression - views suicide as an outcome of an altered decision process, a view supported by preliminary behavioral and imaging data. Thus, the applicant's career goals are to apply advances in the basic neuroscience of decision-making to investigate the brain mechanisms of late-life suicide and, in the future, to identify intervention targets. This approach aims to bridge existing cognitive research on suicidal diathesis with basic and clinical neuroscience. This will be achieved through training in functional magnetic resonance imaging (fMRI; including computational model-based fMRI), in neurobiology of decision-making, and in using neuroscience models to identify intervention targets. An fMRI study of decision-making in elderly depressed suicide attempters focused on disruptions in the processing of value signals in the ventromedial and ventrolateral prefrontal cortex will serve as a vehicle for proposed training. Proposed training and research will be supported by the Advanced Center for Interventions and Services Research for Late-Life Mood Disorders directed by CF. Reynolds (mentor), G.J. Siegle's fMRI lab (co-mentor), and L. Clark's imaging lab (Cambridge, UK). Taking advantage of the neuroimaging resources at the University of Pittsburgh, H.J. Aizenstein will help the applicant to overcome challenges in MR imaging of the aging brain and M.L. Phillips will aid in refining the neural model and developing a life-span perspective on decision-making in suicide. K. Szanto will help the applicant sharpen clinical characterization and manage suicidal risk. An ROl geared towards identifying intervention targets should result from this work.